The zero-crossing analysis of a speech signal in the short-term Method of automatic speaker identification

Downloads

Authors

  • Czesław BASZTURA Institute of Telecommunication and Acoustics, Technical University, Poland
  • Jerzy JURKIEWICZ Institute of Telecommunication and Acoustics, Technical University, Poland

Abstract

The aim of this paper is to investigate the possibility of using the zero-Crossing analysis of a speech signal in a short-term method of speaker identification. Four sets of parameters obtained with the aid of the zero-crossing analysis are presented which can find application in automatic speaker identification. An experiment of speaker identification for 20 male speakers has been performed. The obtained results have confirmed the applicability of the method of zero-crossing analysis for tracking individual features of voices.

References

[1] B. S. ATAL, Automatic recognition of speakers from their voices, Proceedings of IEEE, 64, 4 460-475 (1976).

[2] C. BASZTURA and W. MAJEWSKI, The application of long-term analysis of the zero-crossings of a speech signal in speaker automatic identification, Archives of Acoustics 3, 1 (1978).

[3] R. GUBRYNOWICZ, Problem of the recognition of individual features of voice, Prace IPPT PAN, 28 (1969) (in Polish).